AI Business Matrix - Chaos Tax Solutions

08/05/2026

One in three engineers who respond to major incidents has sought mental health support because of that work.

60% spend at least one full day a week on outage management.

The default response is usually: buy better monitoring.

More sophisticated alerting.

A premium plan.

But I've watched teams do that and come out the other side with the same incident frequency and the same burnout.

The alert arrives faster in the same broken process.

I wrote about what actually changes it - and why the answer is almost never the tool.

06/05/2026

The cost of the incident silence

Your engineering team is heads-down fixing the problem.

Meanwhile, your clients have heard nothing for ninety minutes.

Your customer service team is fielding calls with no information to give.

Your CEO found out from a client, not from you.

The technical team did everything right. The communication was entirely absent.

These two things happen at the same time during every major incident. Most companies only manage one of them.

I wrote about what that costs - and what the structure actually looks like.

During a major IT incident, two conversations happen simultaneously. Most companies only manage one.

01/05/2026

𝐌𝐚𝐧𝐚𝐠𝐞𝐝 𝐒𝐞𝐫𝐯𝐢𝐜𝐞 𝐏𝐫𝐨𝐯𝐢𝐝𝐞𝐫 𝐒𝐋𝐀 𝐓𝐞𝐬𝐭

Your managed service provider has an SLA.

When did you last read it?

Most companies sign it, file it, and don't think about it again - until a major incident hits on a Saturday and the response time they assumed they had turns out not to be what the contract actually says.

I've seen this happen. It's an expensive lesson to learn during a live incident.

Three questions you should be able to answer before your next incident - I covered them here.

Most companies sign their MSP SLA and never read it again - until a live incident exposes the gap between assumed and actual support. Here's what to check before it costs you

29/04/2026

𝐖𝐡𝐲 𝐝𝐨 𝐲𝐨𝐮 𝐧𝐞𝐞𝐝 𝐚 𝐬𝐨𝐫𝐭 𝐨𝐟 𝐈𝐓 𝐠𝐨𝐯𝐞𝐫𝐧𝐚𝐧𝐜𝐞 𝐢𝐧 𝐚 𝐬𝐦𝐚𝐥𝐥 𝐛𝐮𝐬𝐢𝐧𝐞𝐬𝐬?

Nobody gets excited about IT governance.

I get it. It sounds like paperwork, policy documents and audit trails.

But here's something I see regularly: companies spend months evaluating incident management tools without ever defining the process the tool needs to serve.

The tool is sending faster alerts to a team that still doesn't know who does what.

Process first. Tool second. Always.

I wrote about what governance is actually for - and why it matters far more than it sounds.

IT governance isn't just paperwork for large enterprises. For companies of 15–200 people, the right governance practices directly reduce incidents, shorten response times, and prevent the 3 am call

27/04/2026

When key people leave, they don't just take their laptop and their mug.

They take the system knowledge that was never written down.

The reason things are set up the way they are. The workarounds everyone relied on without realising it.

I've watched companies discover this the hard way - during a live incident, when the people who would have known what to do are no longer there.

It's preventable.

But only if you treat knowledge management as a continuous practice, not something you do in a hurry when someone resigns.

When experienced IT staff leave, they take critical system knowledge with them. Learn how knowledge debt compounds incident risk - and what to do about it before the next departure

24/04/2026

Site down. Engineering team working hard.
CEO/Owner finds out from a client. Not from his own IT team.

This happens more often than anyone likes to admit. Not because the technical team isn't doing their job - but because the communication structure was never built.

During a major incident, there are two wars happening at once. Most companies send one person to fight both of them.

I wrote about what that costs - and what actually fixes it.

When a site goes down, most IT teams focus entirely on the fix - and say nothing to clients or the CEO. Here's why that silence costs more than the outage, and what the structure looks like that prevents it

22/04/2026

What should be common between an incident manager and a radio presenter?

I used to present a live radio show.

Then later, I spent 13 years running IT operations.

The skills turned out to be identical.

Communicate clearly when you don't have all the information. Don't let the silence speak for you. Stay calm, because your composure is part of the message.

Most IT teams have never been trained on this. And it shows - every time a client goes quiet after a poorly handled incident.

The tools handle the alert. Humans handle everything that matters after it

Clear IT incident communication under pressure is a skill almost nobody trains for. Here's what 22 years in IT operations - and a career in live radio - taught me about getting it right

20/04/2026

Your best engineer goes on holiday.

Four days in, something breaks.

Nobody else knows the system well enough to fix it quickly.

This isn't a hypothetical.

I've watched it happen.

And it's almost always entirely preventable.

The problem isn't the engineer. It's that the knowledge was never anyone else's to begin with.

Worth a read if you've ever thought "we'd be in trouble if [person] wasn't available."

When your most knowledgeable engineer is sick or on holiday and a major incident hits, how does your team cope? Here's what I've seen happen - and how to prevent it.

17/04/2026

Same incident. Six months later. More expensive than the first time.

I've watched this happen more times than I'd like to admit.
Not because teams are careless - but because post-mortems are treated as a filing exercise rather than an accountability system.

The document gets written.
The actions don't get done.
The same problem comes back.

If your team keeps firefighting the same fires, this is worth a read.
→

Most IT post-mortems are well-written documents. Many of the incidents they cover happen again. Here's why post-mortems fail - and what a process that actually prevents recurrence looks like.

15/04/2026

£11,000 per minute sounds like a big company problem. I understand why.

But in 13 years running IT teams, I've watched smaller companies absorb incident costs that were far larger than they realised - because they only counted the obvious part.

The technical outage. Not the staff hours, the management time, the support calls, the client who quietly started looking elsewhere.

I wrote about what the real number looks like - and where most of it actually comes from.

-> https://aibusinessmatrix.co.uk/post/What-11000-Per-Minute-Actually-Looks

13/04/2026

Here's something I saw over and over again in 14 years of running IT teams.

A company has all the right tools. Everything fires exactly when it should.

And the incident is still a mess.

Because tools tell you something is broken. They don't tell you how to coordinate six teams at once, what to say to your clients, or how to stop your CEO from finding out about the outage from a customer complaint.

That's a human problem. And it's the part most companies haven't solved yet.

I wrote about it here →

Incident Communication failures.